The overall goal of our work is to decrease the infection rate in total knee arthroplasty (TKA) and reduce the treatment morbidity for those patients with TKA that do become infected, thereby benefitting a large and rapidly growing patient population. Sepsis occurs in 1-3% of primary and 3-7% of revision TKA and is by some accounts already the main reason for revision TKA surgery. Treatment of infected TKA often requires prosthetic revision, and the most successful treatment is a costly and morbid two-stage revision surgery that requires specially trained surgeons, has prolonged recovery and creates a rising burden on the health care system. With over 600,000 TKA performed annually in the United States, this complication is a looming public health problem with significant morbidity and mortality. Our immediate goal is to prevent the establishment of infection following TKA. While achieving this goal will reduce the burden of disease associated with infected TKA, our eventual goal is optimize two-stage revision and to produce a robust and reliable one-stage revision for infected TKA, which would dramatically simplify the treatment of infected TKA and reduce both morbidity and related expense. Our overall strategy is organized with two Specific Aims to evaluate two antibiotic delivery methods (free and controlled) with two antibiotics (tobramycin and doxycycline) in our rodent model of infected TKA. Our hypotheses reflect that evidence of infection with a controlled-release antibiotic delivery method will be less than with free release, which will be less than systemic antibiotic or control (no treatment), and that the antibiotics wil differ in their respective responses. Evidence of infection is based on radiographic score, histology (foreign body reaction, inflammation, bony lysis), and quantitative bacteriology. Specific Aims are coordinated to provide standardized data to enable future optimization studies. Although the R21 does not require preliminary work, we include several studies to demonstrate our expertise in infection modeling, local antibiotic delivery, and with the basic and clinical use of the proposed controlled-delivery system. The knowledge we gain will lead to future work evaluating different antibiotic release characteristics and different bacterial strains in combination with parenteral antibiotics.This all leads to our eventual goal of refining a two-stage and developing a reliable, less costly and less morbid one-stage revision surgery. Our multidisciplinary team of material and basic scientists, practicing joint replacement surgeons and orthopaedic infectious disease specialists, along with our strong preliminary work and institutional focus on orthopaedic infection, provide broad background to carefully design and interpret experiments to improve patient outcome.